168 Perkins — Velocity of the Propagation of Magnetism. 



Next by connecting to B, T> b and thus Y b was obtained. Then 

 by connecting A and B in series in such a way that they were 

 in accord, V a+6 , and finally when opposed V a -_ & , were observed. 

 The accent above " a" in the last quantity refers to the necessity 

 of reversing the connections of A (or B), and, strangely enough, 

 the deflection given by any coil was slightly altered on reversal. 

 The'reason for this I was unable to determine, unless, indeed, it 

 is due to inequalities between the positive and negative loops of 

 the commercial current which was used. This seems unlikely, 

 and I purpose to investigate the matter further. This effect 

 was not harmful if properly observed and allowed for. 



With the four readings just referred to it was possible to 

 calculate the angle of lag in two ways that can best be seen in 

 graphic form. 



If the two vectors Y a and Y 6 were in phase, then Y a+b must 

 be equal to their arithmetical sum, otherwise it will be less, and 



P//)Gff/\M OF VECTORS 



this was always found to be the case. The construction can be 

 seen from a glance at the diagram, which gives us <f>, or the 

 angle of lag. Similarly with the two coils opposed, Y a ,„ b will 

 be greater than the arithmetical difference of Y a , and Y b , as 

 shown above. In the results recorded here the latter method 

 was used as being less liable to error, for it involved no large 

 deflection like D a+b and for small values of <j> it is also true from 

 trigonometrical reasons. 



Determinations of cj> were made at intervals up to nine inches 

 from the central coil, and from these values the velocities for 

 the parts in question were calculated. From inspection of the 



curve it is clear that ^ 7 — is not a constant, hence the velocity 



cannot be constant. The velocity at any point in the bar was 

 found from the angle curve by observing the change in lag for 

 a short distance on either side of the point in question. Calling 



